[Rd] Error in glm(..., family=quasi(..., variance=list(...)))
In a glm() call using a quasi() family, one may define a custom variance
function in the form of a "list containing components varfun, validmu,
dev.resids, initialize and name" (quoting the help page for family). In trying
to do so, I run into the following issue that I have not seen discussed
previously:
x <- runif(1000, min=0, max=1)
y <- x + rnorm(1000, mean=0, sd=1)*x^(3/4)
vf <- function(mu) { abs(mu)^(3/4) }
vm <- function(mu) { rep(TRUE, length(mu)) }
dr <- function(y, mu, wt) { (y-mu)^2 }
it <- expression({ n <- rep.int(1, nobs); mustart <- y })
glm(y ~ x, family=quasi(link="identity", variance=list(varfun=vf, validmu=vm,
dev.resids=dr, initialize=it, name="custom")))
This gives "Error in switch(vtemp, constant = { : EXPR must be a length 1
vector" from line 576 in file family.R
(https://github.com/wch/r-source/blob/5a156a0865362bb8381dcd69ac335f5174a4f60c/src/library/stats/R/family.R#L576).
I believe this is due to line 573 "vtemp <- substitute(variance)" and 574 "if
(!is.character(vtemp)) vtemp <- deparse(vtemp)" where vtemp becomes a length 2
character vector because, by default, deparse() breaks lines at width.cutoff =
60L characters. In stepping through quasi() during debug, setting vtemp <-
(vtemp, collapse=" ") on line 576 avoids the error.
A workaround from https://tolstoy.newcastle.edu.au/R/help/05/06/6795.html
appears to be to define one's own complete quasi2() function with the desired
variance function pre-stored.
Is this known/expected, or should I file a bug?
Many thanks and best regards
Daniel
> sessionInfo()
R version 3.6.0 (2019-04-26)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 14393)
Matrix products: default
locale:
[1] LC_COLLATE=German_Germany.1252 LC_CTYPE=German_Germany.1252
[3] LC_MONETARY=German_Germany.1252 LC_NUMERIC=C
[5] LC_TIME=German_Germany.1252
attached base packages:
[1] stats graphics grDevices utils datasets methods base
loaded via a namespace (and not attached):
[1] compiler_3.6.0
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
[Rd] Problem with a regular expression.
The issue seems related to R bug report 15012:
https://bugs.r-project.org/bugzilla3/show_bug.cgi?id=15012
As mentioned in the comments there, a pull request to the TRE library has
recently been made, but I don't know about its status.
Daniel
Von: R-devel im Auftrag von Tomas Kalibera
Gesendet: Donnerstag, 17. August 2017 10:14
An: r-devel@r-project.org
Betreff: Re: [Rd] Problem with a regular expression.
The problem is in TRE library, in regcomp, while compiling the regular
expression.
This is enough to trigger in R (to do this without re-boot: ulimit -v
50 ):
> strsplit("", ")")
To repeat in TRE, one can build TRE from sources and run
> ./src/agrep ")" README.md
Tomas
On 08/17/2017 09:45 AM, Moshe Olshansky via R-devel wrote:
> I tried this on a Linux (Ubuntu) server invoking R from the command line and
> the result was the same, except that I could kill the R session from another
> terminal window.
>
>
>From: Rui Barradas
> To: Chris Triggs ; "r-devel@r-project.org"
>
> Cc: Thomas Lumley
> Sent: Thursday, 17 August 2017, 17:26
> Subject: Re: [Rd] Problem with a regular expression.
>
> Hello,
>
> This seems to be serious.
> RGui.exe, fresh session. I've clicked File > New Script and wrote
>
> Oldterm <- c("A", "B", "A", "*", "B")
> strsplit(Oldterm, ")" )
>
> Ran each instruction at a time with Ctrl+r and with the strsplit call
> the system froze.
>
> Ctrl+Alt+Del didn't work, I had to go for the power switch button.
>
> sessionInfo()
> R version 3.4.1 (2017-06-30)
> Platform: x86_64-w64-mingw32/x64 (64-bit)
> Running under: Windows 7 x64 (build 7601) Service Pack 1
>
> Matrix products: default
>
> locale:
> [1] LC_COLLATE=Portuguese_Portugal.1252
> LC_CTYPE=Portuguese_Portugal.1252
> [3] LC_MONETARY=Portuguese_Portugal.1252 LC_NUMERIC=C
>
> [5] LC_TIME=Portuguese_Portugal.1252
>
> attached base packages:
> [1] statsgraphics grDevices utilsdatasets methods base
>
> loaded via a namespace (and not attached):
> [1] compiler_3.4.1
>
>
> Rui Barradas
>
> Em 16-08-2017 23:31, Chris Triggs escreveu:
>> Hi...
>>
>> I have come upon a problem with a regular expression which causes base-R to
>> freeze. I have reproduced the phenomenon on several machines running R
>> under Windows 10, and also under OSX on different Apple MACs.
>>
>> The minimal example is:-
>> Oldterm is a vector of characters, e.g. "A", "B", "A", "*", "B"
>> The regular expression is ")"
>>
>> The call which freezes R is
>> strsplit(Oldterm, ")" )
>>
>> Thomas - after he had reproduced the problem - suggested that I submit it to
>> r-devel.
>>
>> Best wishes
>> Chris Triggs
>>
>>
>> [[alternative HTML version deleted]]
>>
>> __
>> R-devel@r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-devel
>>
> __
> R-devel@r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
>
>
>
> [[alternative HTML version deleted]]
>
> __
> R-devel@r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] undocumented 'offset' argument in src/library/grDevices/man/adjustcolor.Rd
One might say that adjustcolor() performs affine color transformation. Argument
'offset' defines the shift in color space after the linear transformation
defined by argument 'transform'. In other words, 'offset' represents additive
color mixing (in the sense of adding lights, not paints) after linear
transformation.
> red <- rgb(1, 0, 0)
> green <- c(0, 0.9, 0, 1)
> adjustcolor(red, 1, 0.9, 0.9, 0.9, offset=green)
[1] "#E6E600FF"
("red + green = yellow")
However, the implementation of adjustcolor() is a bit inconsistent with this
description as it does not force 'offset' to be a color by using
col2rgb(offset, alpha=TRUE)/255. The help page already gives an example for
using 'offset' for the purpose of adding white light.
Von: R-devel im Auftrag von Martin Maechler
Gesendet: Mittwoch, 2. Dezember 2020 09:45:27
An: Ben Bolker
Cc: r-devel@r-project.org
Betreff: Re: [Rd] undocumented 'offset' argument in
src/library/grDevices/man/adjustcolor.Rd
> Ben Bolker
> on Mon, 30 Nov 2020 16:33:23 -0500 writes:
> The 'offset' argument description is blank ...
> maybe 'additive adjustment to each of the (red, green, blue, alpha)
> values defining the colors, after adjustment by the corresponding
> \code{.f} factor' ...?
Thank you, Ben. I'm using adjustcolor() very often, and I had
also wondered about the documentation "FIXME" there.
As I've not been a real expert in color space transformations, I
hadn't dared to describe it myself, but your description above
is clearly better than the emptyness ...
Maybe someone can add something, or contribute some example(s)
on how using offset != 0 makes sense ?
> This is the relevant code:
> x <- col2rgb(col, alpha = TRUE)/255
> x[] <- pmax(0, pmin(1,
> transform %*% x +
> matrix(offset, nrow = 4L, ncol = ncol(x
> rgb(x[1L,], x[2L,], x[3L,], x[4L,])
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
